Diagnosis of dementia of the Alzheimer type (DAT) presents a major challenge to clinical medicine because the early cognitive changes in DAT are difficult to distinguish from those accompanying normal adult aging. Positron emission tomography (PET) studies have shown, however, that the topography of cerebral metabolism is altered in the early stages at Alzheimer's disease (AD). These studies point to selective, corticocortical disconnections of association areas in AD, in particular the posterior parietal, temporal, and frontal cortices. These areas play a major role in the control of attentional functions. Thus, the corticocortical pathways affected by AD are also substantially involved with cortical systems of attention. Recent studies indicate that attentional impairment can be detected in mildly demented subjects and may precede deficits in all other nonmemory functions affected by AD. A neuroanatomical model of attention in DAT is proposed, postulating dysfunction of separate but interacting corticocortical circuits associated with different forms of visual selective attention. Circuits involving posterior parietal, extrastriate, and association temporal areas are postulated to mediate location-based, feature-based, and object-based selective attention, respectively. A series of six studies of DAT patients, healthy, age-matched controls, and normal young and older adults is proposed. Measures of attention and resting cerebral metabolism will be available in all studies; PET imaging of attention-related cerebral blood flow will be examined in one study. Two broad aspects of visual selective attention in early DAT will be examined, location-based and feature/object-based selection, using well-established information-processing tests of covert visuospatial attention and visual search. Normal, age-related changes in these two aspects of visual selective attention will also be investigated. Location-based selective attention in aging and DAT will be investigated in the first three experiments, including a longitudinal study examining changes over a three-year period in initially mild DAT subjects. The last three experiments will examine feature-based and object-based selective attention and their interactions with visuospatial attention in normal aging and DAT. The neuroanatomical model of attention will be evaluated by (i) using attention tasks that are known to depend upon the functional integrity of cortical areas represented in the model; and (ii) using PET imaging of both resting cerebral metabolism and task-related cerebral blood flow to index dysfunctions in the postulated neocortical areas and their relation to attentional performance in mild DAT. The proposed research is significant because changes in attention in the early stages of DAT have not been studied extensively. Studies of early changes in cognitive abilities (attention) and physiological functioning (cerebral metabolism and blood flow) and their interrelationships are essential for developing reliable markers for early diagnosis of DAT. The research will also improve understanding of the normal aging of attention. The unique contribution of the research is that brain metabolic changes will be examined in relation to attentional changes associated with both normal aging and DAT.